Glossary of Terms

.exe file - An “.exe” file, short for “executable,” is a common file extension used in Windows operating systems and some other computing environments. An executable file contains a program or application that can be run or executed by a computer’s operating system.

.usd - Universal Scene Description (Binary) This is the standard binary or ASCII file format for USD. It stores the 3D scene and asset data in a compact, binary form, making it efficient for storage and processing.

.usda - Universal Scene Description (ASCII) This format stores USD data in a human-readable, ASCII text format. It’s primarily used for debugging and as a reference because it’s easier for humans to read and modify. However, it’s less efficient in terms of file size and loading speed compared to the binary format.

.usdc - Universal Scene Description (Crate) This is a binary format for USD, but it’s optimized for high-performance data storage and retrieval. .usdc files are typically used as the primary format for asset storage and production pipelines, as they offer faster loading and saving times compared to .usd files.

.usdz - Universal Scene Description (ZIP) A compressed container file in the ZIP structure that can contain both geometry and texture information.


Anisotropic Roughness - Refers to a property that describes surface roughness in a way that varies based on the direction of measurement. Unlike isotropic roughness, which is uniform in all directions, anisotropic roughness implies that the micro-surface irregularities on an object’s surface have a directional preference.

Anisotropy - Refers to a property of materials or surfaces that causes them to exhibit different reflective or shading characteristics in different directions. Anisotropic materials have a direction-dependent behavior, meaning they can appear shinier or more reflective in one direction while exhibiting different properties in other directions.


Barycentric Coordinates - A set of coordinates used to describe the position of a point within a triangle or other convex polygon. These coordinates are defined relative to the vertices of the polygon and are useful for various operations in graphics rendering, including interpolation and texture mapping. Barycentric coordinates are represented as a set of weights for each vertex of the polygon.

Bitangent - A vector that is perpendicular to both the surface normal and the tangent vector of a 3D surface at a particular point. The bitangent vector is typically used in advanced shading techniques, such as normal mapping and bump mapping, to complete a local coordinate system known as the tangent space.

Blue Noise - Refers to a type of noise pattern that has unique properties, making it particularly useful for various applications, including texture mapping, sampling, and anti-aliasing. Blue noise is characterized by a distribution of points or values in such a way that they are more evenly spaced and have a more perceptually uniform distribution of energy in the high-frequency spectrum, especially in the blue part of the spectrum.


Composite Output - Refers to the final image that is generated by combining and blending various graphical elements or layers together. This process typically involves taking multiple rendered images, often with transparency information, and compositing them into a single cohesive image that represents the final scene as it will be displayed to the viewer.

Cone Radius - The path of a ray of light as it extends from the camera into the scene and helps determine which objects or surfaces in the 3D environment the ray intersects with.


DDS File - DDS stands for “DirectDraw Surface,” and it is a file format commonly used in computer graphics and game development. DDS files are specifically designed for storing and efficiently accessing texture and image data.

Diffuse Albedo - Refers to the inherent color or reflectance of a surface when it interacts with and scatters incoming light uniformly in all directions. It represents the base color or the color of a surface under diffuse lighting conditions, meaning when there are no specular highlights or reflections.

Disocclusion - refers to the process of determining what is visible and what is not in a 3D scene from a given viewpoint. It primarily relates to the handling of objects, surfaces, or portions of objects that were previously hidden or occluded by other objects but have become visible due to changes in the camera’s position or orientation.

DLSS - NVIDIA DLSS (Deep Learning Super Sampling) is a neural graphics technology that multiplies performance using AI to create entirely new frames and display higher resolution through image reconstruction—all while delivering best-in-class image quality and responsiveness.


Emissive Radiance - Refers to the radiant energy emitted or radiated from a surface or object in a 3D scene. It represents the light or color that a surface emits as opposed to reflecting or scattering light like most materials.

Exposure Histogram - A graphical representation that provides a visual summary of the distribution of pixel brightness or luminance values in an image. It shows how many pixels fall into different brightness or exposure levels, typically displayed as a histogram chart.


Froxel - a portmanteau of “fragment” and “voxel” used in computer graphics. It represents a small 3D volume element or pixel-sized voxel within a three-dimensional space. Froxels are typically used in volume rendering and ray tracing techniques to sample and process data within a 3D volume, similar to how pixels sample a 2D image.


HitT - Hit-Testing (hit detection, picking, or pick correlation) is the process of determining whether a user-controlled cursor (such as a mouse cursor or touch-point on a touch-screen interface) intersects a given shape, line, or curve drawn on the screen.


Inf/Nan Check - Refers to a process or technique used to identify and handle numerical values that are either infinite (Inf) or not-a-number (NaN) during rendering or computation.

Interpolated Normal- A normal vector calculated for a specific point on a 3D surface by interpolating or blending the normals of nearby vertices. These interpolated normals are used to determine how light interacts with the surface and are crucial for achieving smooth shading and realistic lighting effects.

Isotropic Roughness - Refers to a property that describes the degree of micro-surface irregularities or roughness on a 3D object’s surface in a uniform and non-directional manner. This roughness affects how light scatters and interacts with the surface, leading to diffuse reflections.


Local Tonemapper Luminance Output - Refers to the output of a local tonemapping process that adjusts the brightness and contrast of an image on a pixel-by-pixel basis. This adjustment is based on the luminance or brightness values of the pixels in the image.


Material Type - refers to a classification or categorization of the physical properties and visual characteristics of surfaces or materials used in 3D scenes. Material types are used to describe how light interacts with a particular surface and how it should be shaded and rendered. Common material types might include: Diffuse Materials, Specular Materials, Translucent Materials, and Emissive Materials.


Normals - Perpendicular vectors on the surfaces of 3D objects. They define the orientation of surfaces, play a key role in lighting calculations, and enable smooth shading by interpolating across polygon surfaces. Normals are fundamental for simulating how light interacts with objects and achieving realistic lighting and shading effects in 3D scenes.

NRD - NVIDIA Real-Time Denoisers (NRD) a spatio-temporal, API-agnostic denoising library that’s designed to work with low ray-per-pixel signals. It uses input signals and environmental conditions to deliver results comparable to ground-truth images.


Octahedron-normal vectors - to encode normals by projecting them on an octahedron, folding it and placing it on one square to give it uniform properties for value distribution and low encoding and decoding costs.

Opacity - Refers to the degree to which an object or part of an object is transparent or allows light to pass through. It is a fundamental property used to control the visibility and transparency of 3D objects and their components within a rendered scene.


Pixel Checkerboard - A technique used to analyze and visualize the distribution of pixel shading workloads across the screen or image. It involves rendering a checkerboard pattern over a scene, where each square of the checkerboard represents a pixel. The color of each square can indicate the complexity or computational workload of the corresponding pixel.

Primary Depth - Refers to the depth information associated with the primary rays in a ray tracing pipeline.

Primary Ray Bounces - Refers to the first set of rays cast from the camera or viewer into a 3D scene during the ray tracing process. These primary rays are used to determine which objects or surfaces in the scene are visible from the camera’s perspective.

Primary Specular Albedo - Refers to the albedo or reflectance value that represents the color and intensity of the specular reflections on a surface when using ray tracing or other rendering techniques. It specifically relates to how a surface reflects light from direct light sources, such as point lights or directional lights.

Primitive Index - Refers to a unique identifier or index associated with a primitive in a rendering or graphics pipeline. Primitives are fundamental geometric shapes or elements used in computer graphics to construct more complex scenes and objects. These primitives can include points, lines, triangles, and more. Each primitive is assigned a unique index that allows the GPU to process them individually or in specific groups as required by the rendering algorithm.


ReBLUR - A denoiser based on the idea of self-stabilizing, recurrent blurring. It’s designed to work with diffuse and specular signals generated with low ray budgets. In fact, ReBLUR supports checkerboard rendering, producing reasonable results when casting just half a ray per pixel.

ReLAX - a variant of SVGF optimized for denoising ray-traced specular and diffuse signals generated by NVIDIA RTX™ Direct Illumination (RTXDI). ReLAX offers substantial improvements to image quality and performance over stock SVGF. Not only does ReLAX preserve lighting details produced by massive RTXDI light counts, it also yields better temporal stability and remains responsive to changing lighting conditions.

ReSTIR Direct Illumination - ReSTIR or spatiotemporal reservoir resampling samples one-bounce direct lighting from many lights without needing to maintain complex data. ReSTIR DI samples all primary lighting and shadows in the screen space 65x faster than the previous state of the art solution (RIS or resampled importance sampling). This screen space light sampling solution is capable of virtually unlimited lights with a few (1-4) rays per pixel.

ReSTIR Global Illumination - ReSTIR GI resamples multi-bounce indirect lighting paths. At a single sample per pixel every frame, this solution achieves a mean-square error (MSE) improvement greater than 10x. In conjunction with a denoiser, this offers high quality path tracing at real-time frame rates.

RTX - RTX represents real-time ray tracing, where the calculations required for ray tracing are performed in real-time, enabling lifelike graphics and dynamic lighting effects in video games and other applications.

RTXDI - (RTX Direct Illumination) Generates millions of fully ray traced dynamic lights creating photorealistic lighting of night and indoor scenes that require computing shadows from 100,000s to millions of area lights. No more baking, no more hero lights. Unlock unrestrained creativity even with limited ray-per-pixel counts. When integrated with RTXGI and NVIDIA Real-Time Denoiser (NRD), scenes benefit from breathtaking and scalable ray-traced illumination and crisp denoised images, regardless of whether the environment is indoor or outdoor, in the day or night.

RTXGI - {RTX Global Illumination} Multi-bounce indirect light without bake times, light leaks, or expensive per-frame costs. RTX Global Illumination (RTXGI) is a scalable solution that powers infinite bounce lighting in real time, even with strict frame budgets. Accelerate content creation to the speed of light with real-time in-engine lighting updates, and enjoy broad hardware support on all DirectX Raytracing (DXR)-enabled GPUs. RTXGI was built to be paired with RTX Direct Illumination (RTXDI) to create fully ray-traced scenes with an unrestrained count of dynamic light sources.


Screen-Space Motion Vector - (SSMV) is a data representation that captures the motion of objects or pixels between consecutive frames within the screen space. SSMVs are used in various rendering techniques, such as motion blur and temporal anti-aliasing, to simulate realistic motion effects.

Secondary Ray Bounces - Refers to the rays that are cast after the primary rays during the ray tracing process. Primary rays are initially shot from the camera or viewer into the 3D scene to determine which objects or surfaces are visible. When these primary rays hit a reflective or refractive surface, secondary rays are generated to simulate additional lighting effects, such as reflections and refractions.

Secondary Specular Albedo - Refers to the albedo or reflectance value that represents the color and intensity of the secondary, or indirect, specular reflections on a surface when using ray tracing or other advanced rendering techniques.

Shading Normal - Often referred to as a “Geometric Normal,” is a vector that represents the orientation or facing direction of a surface at a particular point on a 3D object. This vector is used in shading calculations to determine how light interacts with the surface and how the surface should be illuminated or shaded.

SIGMA - SIGMA is a fast shadow denoiser. It supports shadows from any type of light sources, like the sun and local lights. SIGMA relies more on physically based spatial filtering than temporal filtering, offering minimal temporal lag.

Stochastic Texture Filtering (STF) - often referred to as stochastic texture synthesis or stochastic texture generation, is a technique used in computer graphics and computer vision to create realistic and natural-looking textures. The term “stochastic” refers to randomness or unpredictability, and in this context, it involves introducing controlled randomness into the generation of textures to make them appear more natural and less repetitive.


Tangent- Refers to a vector that lies in the plane of a 3D surface and is perpendicular to the surface’s normal vector. Tangent vectors are used in techniques like normal mapping and bump mapping to simulate fine surface details and enhance the realism.

Texture Coordinates - Texture coordinates specify which part of a 2D texture map should be applied to each point on the 3D model’s surface. This mapping allows for realistic and detailed texturing of 3D objects, enabling the application of textures like color, bump maps, normal maps, and more. By specifying texture coordinates for each vertex, the graphics hardware can interpolate and apply the corresponding texture data smoothly across the entire surface, creating the illusion of complex surface properties and details on the 3D object when rendered.

Thin Film Thickness - Refers to the measurement of the thickness of a thin film or layer of material that is applied to a surface. This concept is often used in computer graphics and rendering to simulate the interaction of light with thin layers of materials, such as oil films on water or soap bubbles.

Tonemapping - A technique of converting high dynamic range (HDR) images or scenes into low dynamic range (LDR) images that can be displayed while preserving as much visual detail and realism as possible.

Triangle Normal- These normals help determine how light rays are reflected or refracted off the triangle’s surface, affecting the appearance of the triangle in the rendered image.


UDIM- Stands for U DIMension is based on a tile system where each tile is a different texture in the overall UDIM texture array where each tile consists of its own UV space (0-1, 1-2, 2-3) and has its own image assigned to that tile.


Vertex Color - Refers to color information associated with individual vertices (corner points) of a 3D model or object. Each vertex can have a specific color value assigned to it, which is typically represented as a combination of red, green, and blue (RGB) values.

Virtual Motion Vector - The motion of objects between frames. These motion vectors help reduce data redundancy and improve compression efficiency by describing how objects move from one frame to another.

Need to leave feedback about the RTX Remix Documentation? Click here